However, I quickly realized that such a full immersion environment simply wasn't feasible, nor enjoyable for me - at least, not in the way Khatz has it all laid out.
I also realized how much he over exaggerates in order to get his points across, and I think this is something a lot of people don't recognize. I don't believe that he literally wants his readers to ditch their friends in favor of Japanese speaking ones, nor do I believe that he literally means we should all get rid of our non-Japanese music, or deck out our living quarters in traditional Japanese decor, all for the sake of immersion.
These are extremes that may have worked for him, personally, but things I'd never even consider touching with a ten foot pole. I love learning this language and I'm very comfortable at the pace I'm going, and to sacrifice things so dear to me for the sake of learning a little faster is beyond ridiculous.
Besides that, living in Japanese simply begins to wear on me, and when I want to do something very specific and technical on my operating system, for example, I'd really rather not stumble around in the dark and risk doing something stupid like accidentally formatting my hard drive (yes, now I'm exaggerating to get my point across). It's not fun, it's not productive - at least, not at the point I'm at currently. I can certainly see myself pushing immersion to this level a year from now, when I can really get a lot more out of everything I've learned up to that point, but if the immersion isn't enjoyable for me, I don't feel as though I'm missing much by forgoing it.
Which leads me to another criticism - Khatz does a decent job of explaining his whole system, but really hurls you out of the nest straight into the great beyond once it comes time to really start things. Learning kanji before kana? I can see the benefits for a beginning learner who is absolutely, unfalteringly dead serious about learning Japanese, but for the vast majority of those interested in learning the language, this approach is ridiculous. Considering how much of the important fundamentals one could soak up in a short period of time from a good beginner's text book or a well instructed class, it boggles my mind to recommend leaping straight into kanji before learning the kana.
And that leads me to my next point: the propaganda that all language classes are a horrible, despicable, draconian thing to be avoided at all costs. I'm truly sorry if your language class experiences were that bad, but the vast majority of my language teachers were nothing less than wonderful. There's nothing inherently bad about language classes; it's absolutely about the teacher and how they teach the language. Assuming a class isn't too infrequent, I can think of few better ways to build a solid foundation for any language. Of course you're not going to become fluent from classes alone, but this is common sense.
Another point which many AJATTeers seem to get caught up on (which is only their own fault) is Khatzu's "10,000 sentences" model which, while a pretty darn good guideline for all intents and purposes, should by no means be the ultimate goal of the learner. All too often do I see learners trying to import huge quantities of sentences via pre-constructed SRS decks for the sake of reaching this figure of 10,000 faster, but it just doesn't work this way.
Quality of sentences will vary dramatically depending on source, and the method by which the learner is studying said sentences will also have a significant impact (focusing on one word from the sentence? the entire sentence, strictly? going by production from kana to kanji? etc).
Khatz mentions quality of sentences very frequently - if you feel that a sentence sucks, in any way, shape or form, it's destined for deletion. I believe he cites a near 50% deletion rate of his sentences, in fact, which I can definitely believe as I do my fair share of deleting with prejudice, myself. Imagine going through 10,000 without ever deleting one - think of all the garbage cards you'd have to suffer through in the process. On the other hand, assuming you're deleting 50% of these suckers, you're more likely going to go through 15,000 or more before you reach the 10k goal (in theory anyway).
By now, I must sound like I'm an anti-AJATT, purist pundit, but quite the contrary as I've adopted and practice plenty from this system (and indeed, philosophy).
As I've said, I have massive amounts of respect for Khatzumoto himself and AJATT in general. As with all teachers, though (see also: James Heisig, another invaluable 先生 of mine), I believe it's important for individual, mental (and even spiritual) growth for the student to regularly question and challenge, and to keep a critical and open mind at all times.
Presently, I'm progressing at a pace comfortable to me without constant immersion, and I never thought I'd have made it this far in well under a year. I've set goals (big and small) throughout the year which I know I'll obtain, and it's a great feeling. I'm playing through my favorite RPGs in Japanese, I'm reading my favorite manga in Japanese, I'm watching slews of dramas and movies and anime in Japanese and every single day I move a few steps closer to my ultimate goal of fluency and literacy - all of this without full immersion, without worrying about collecting 10,000 sentences, without abandoning all of life's simple, English-language pleasures. Let's call it, All Japanese Some of the Time.
Now if you'll pardon me - it's back to reviewing sentences while rocking out to Iron Maiden, a band I can only wish was in Japanese.